“That means,” Colonel Aureliano Buendía said, smiling when the reading was over, “that all we’re fighting for is power.”
“They’re tactical changes,” one of the delegates replied. “Right now the main thing is to broaden the popular base of the war. Then we’ll have another look.”
One of Colonel Aureliano Buendía’s political advisers hastened to intervene.
“It’s a contradiction,” he said. “If these changes are good, it means that the Conservative regime is good. If we succeed in broadening the popular base of the war with them, as you people say, it means that the regime has a broad popular base. It means, in short, that for almost twenty years we’ve been fighting against the sentiments of the nation.”
He was going to go on, but Colonel Aureliano Buendía stopped him with a signal.
“Don’t waste your time, doctor,” he said. “The important thing is that from now on we’ll be fighting only for power.” Still smiling, he took the documents the delegates gave him and made ready to sign them.”
—Gabriel García Márquez, One Hundred Years of Solitude
“From now on we’ll be fighting only for power.” Colonel Buendía’s smiling resignation to the pursuit of power for its own sake, after decades spent fighting for principle, neatly encapsulates the trajectory of many revolutions, and not only political ones. Silicon Valley, once the engine room of a genuine technological uprising, has followed a strikingly similar arc. What began as a gathering place for engineers and idealists has calcified into an enclave of entrenched capital, now less concerned with innovation than with strategic dominance and the systematic elimination of rivals.
The garage where Hewlett and Packard started their eponymous company is now marked with a placard, “The birthplace of Silicon Valley”. Hewlett-Packard built its reputation not just on its marvellous output of products, including what is generally acknowledged to be the first personal computer, but also on a kind of engineering integrity. Intel’s Moore didn’t just describe a law of exponential progress; he embodied an era of technological purpose. Intel and its rivals in the semiconductor industry were cutthroat competitors, but they were driven by a mission to create excellent microprocessors. The early internet was full of these figures – idealists with a garage and a dream. There was, briefly, a sense that this might all go differently. The optimism was real: Google’s early unofficial motto, coined by Sergey Brin, was literally “Don’t be evil”. It was was more than a line; it was a moral thesis statement. The company’s clean, ad-free interface, its commitment to organizing the world’s information, and its apparently principled stance against intrusive monetization suggested a new kind of corporation. One that meant well. Who didn’t love Google back then?
As for Facebook, it was never quite Utopian, but it was innocuous enough. A college app, fun and forgettable. Harmless. The idea that, little more than a decade later, it would have a policy division, a government relations office, and an operational footprint that spanned from Washington to Myanmar would have seemed absurd. And yet, here we are.
What changed? Everything- and also nothing. The core product was always attention. But somewhere along the way, it began to tap into something deeper than the flicker of eyeballs, the instant of attention. It accessed belief.
The shift wasn’t immediate. Even as late as 2016, Mark Zuckerberg was apparently genuinely confused by the thing he’d built. According to Careless People by Sarah Wynn Williams, it took a ten-hour transatlantic flight for his team to convince him that Facebook had quite literally handed the U.S. presidential election to Donald Trump, executing what remains perhaps the most surgically effective ad campaign in political history.
“The Social Network” became a psychological battleground.
The evolution of Facebook and platforms like it is not merely a matter of reach, but of resonance. They didn’t just enable communication; they amplified some of the worst tendencies in human psychology; they amplified human traits that were better in small doses.
What we’ve created is not propaganda in the traditional sense. It is bespoke propaganda: tailored not for the masses, but for the individual. It is possible to create a reality that is optimized by the best technology in the world to confirm one’s beliefs and bypass discourse entirely.
This was not the plan when Zuckerberg started Facebook. For a very long time seemed to genuinely believe that he was merely “connecting the world”. But the plan with Zuckerberg, as with Buendía, was quietly set aside. What matters now is not connection or truth or progress. What matters is power.
Social media platforms, gig economy apps, and even once-revolutionary tools like Google Search are no longer at the frontier of technology. That mantle now belongs to artificial intelligence (AI).
AI is the technology that promises everything: efficiency, insight, transformation. But unlike earlier eras of technological disruption, this one is not being driven by idealistic upstarts in garages. It is being deployed almost exclusively by incumbents. In the consumer space, Amazon and Walmart are masticating AI – chewing it up and working it into their business models -but they are not in the business of solving grand human problems. They are in the business of achieving logistical supremacy.
The question for these players is not What can AI do? but What fraction of market share can it defend? Or more bluntly: How can it help the biggest players get even bigger, even faster?
These companies already dominate the online consumer market. What is the goal now? To strip us of a few more dollars, a bit more efficiently? To add a few extra seconds to our browsing time, scrolling through dynamically ranked options? To reduce purchasing friction to the point where it becomes a negative coefficient – where you don’t just want the product, but feel vaguely penalized for not buying it?
This isn’t transformation. It’s machine intelligence not for solving the difficult, messy, painful human problems, but at smoothing the path of least resistance toward continued consumption. One cannot help but think that AI is being shaped, deployed and used not by those who need it most, but by those who already have everything. The goal is not enhancement, but entrenchment.
And so the cycle of Colonel Buendia repeats: idealism, compromise, consolidation. There is concrete evidence that like the social media revolution that preceded it, the AI revolution may cause more harm than good. The early AI pioneers, many of them brilliant and well-intentioned, likely envisioned better futures. Even some of their most prominent advocates, such as Geoffrey Hinton and Elon Musk, have called for pause and careful reflection, warning that without caution we risk harming ourselves.
The boy kings of Silicon Valley already possess the riches and wherewithal to fulfill every material desire. However, to be king isn’t a question only of wealth; to remain king one must maintain dominance. This demands ruthlessness and loyalty not to superior products or the service of humanity – regardless of how damaging those products may be – but to growth, market share, and attention. The desire to cling to power, to cling to the throne, is so formidable that in Silicon Valley, to pause and reconsider is effectively to abdicate. Once crowned, it takes an exceptional character to voluntarily descend to commoner status on principle. Ultimately, what counts is not whether the product improves lives but whether it scales—even if the cost is self-destruction.
What’s remarkable is not that the tech revolution betrayed its ideals. That’s expected. What’s remarkable is how quickly, and how smoothly, the transition from principle to power has become protocol.
The question that hovered over Facebook in 2007 now hovers over AI in 2025: What is this for?
Painfully, I think the answer is “power”. Once they became kings, they were fighting only for power.